In conventional numerical control devices, a desired machining is realized by combining machining processes accompanying a plurality of movements of a tool. For example, in a case of performing a machining of forming many holes, the machining is performed by combining a positioning process on an X-Y plane, a boring process of lowering along a Z axis, a process of restoring (retracting) to the positioning plane by rising along the Z axis, and so on. In practice, positioning accuracy required in the respective processes differ. However, usually, since an in-position width is the same in each of the processes, the in-position width is unified to an in-position width of a process in which a high positioning accuracy is required. As a result, this positioning requires much time and hence results in a cause of elongating a machining time.
Methods for solving such problem have been proposed. For example, a high-speed boring method is disclosed in patent document 1. According to the method of the patent document 1, a machining time is intended to be shortened by providing an in-position width at the bottom of a hole, and an in-position width for positioning and an in-position width for retracting set so as to have a value larger than the in-position width at the bottom of a hole.